DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced features, such as DeepPot-SE, attention-based and hybrid descriptors, the ability to fit tensile properties, type embedding, model deviation, DP-range correction, DP long range, graphics processing unit support for customized operators, model compression, non-von Neumann molecular dynamics, and improved usability, including documentation, compiled binary packages, graphical user interfaces, and application programming interfaces. This article presents an overview of the current major version of the DeePMD-kit package, highlighting its features and technical details. Additionally, this article presents a comprehensive procedure for conducting molecular dynamics as a representative application, benchmarks the accuracy and efficiency of different models, and discusses ongoing developments.

A recently discovered two-dimensional (2D) layered material phosphorene has attracted considerable interest as a promising p-type semiconducting material. In this work, thermal conductivity (κ) of monolayer phosphorene is calculated using large-scale classical non-equilibrium molecular dynamics (NEMD) simulations. The predicted thermal conductivities for infinite length armchair and zigzag phosphorene sheets are 63.6^+3.9 _−3.9 and 110.7^+1.75 _−1.75 W m⁻¹ K⁻¹ respectively. The strong anisotropic thermal transport is attributed to the distinct atomic structures at altered chiral directions and direction-dependent group velocities. Thermal conductivities of 2D graphene sheets with the same dimensions are also computed for comparison. The extrapolated κ of the 2D graphene sheet are 1008.5^+37.6 _−37.6 and 1086.9^+59.1 _−59.1 W m⁻¹ K⁻¹ in the armchair and zigzag directions, respectively, which are an order of magnitude higher than those of phosphorene. The overall and decomposed phonon density of states (PDOS) are calculated in both structures to elucidate their thermal conductivity differences. In comparison with graphene, the vibrational frequencies that can be excited in phosphorene are severely limited. The temperature effect on the thermal conductivity of phosphorene and graphene sheets is investigated, which reveals a monotonic decreasing trend for both structures.

As the dimensions of nanocircuits and nanoelectronics shrink, thermal energies are being generated in more confined spaces, making it extremely important and urgent to explore for efficient heat dissipation pathways. In this work, the phonon energy transport across graphene and hexagonal boron-nitride (h-BN) interface is studied using classic molecular dynamics simulations. Effects of temperature, interatomic bond strength, heat flux direction, and functionalization on interfacial thermal transport are investigated. It is found out that by hydrogenating graphene in the hybrid structure, the interfacial thermal resistance (R) between graphene and h-BN can be reduced by 76.3%, indicating an effective approach to manipulate the interfacial thermal transport. Improved in-plane/out-of-plane phonon couplings and broadened phonon channels are observed in the hydrogenated graphene system by analyzing its phonon power spectra. The reported R results monotonically decrease with temperature and interatomic bond strengths. No thermal rectification phenomenon is observed in this interfacial thermal transport. Results reported in this work give the fundamental knowledge on graphene and h-BN thermal transport and provide rational guidelines for next generation thermal interface material designs.